Visual function depends on the remarkable integration of a number of interrelated subprocesses such as anticipation, stimulus recognition, and stimulus discrimination. The brain is able to orchestrate these elementary processes by coordinating the activities of diverse neural structures in the face of continuously varying processing demands. The question of how this coordination operates is central to understanding the neural basis of visual function. This proposal aims to investigate the coordinated activity of distributed neuronal ensembles in the cerebral cortex of non-human primates performing a visual pattern discrimination task. It is motivated by theoretical considerations suggesting that the large-scale functional coordination of interacting neuronal ensembles is an essential component of visual function (Bressler 1995, 1996; Bressler & Kelso 2001). A combined approach of experimental work and methods development is proposed to investigate the operations of large-scale cortical networks underlying visual function. Specific hypotheses concerning the dynamics of large-scale cortical coordination will be tested by analysis of local field potentials recorded from indwelling electrodes in the cerebral cortex. Access to the rapidly changing dynamics of field potential interdependency will be possible through the use of the adaptive multivariate autoregressive methodology developed under our previous R03 project. This methodology consists of a number of interrelated techniques that can characterize the multiple, complex interactions of large-scale distributed cortical networks in a very short time frame (Ding et al. 2000). Included are methods to derive multi-site interaction patterns, analyze the dependencies of functional relations on particular group of neurons, and measure causal influences between cortical areas. Our methodological development will continue by constructing a comprehensive framework for the optimal utilization of existing techniques, and by developing more powerful measures of network function that deal with problems of nonlinearity and nonstationarity. This work is expected to: (1) produce new insights into the coordination dynamics of large-scale cortical networks in vision; and (2) make available new digital signal processing tools for the investigation of large-scale neural systems underlying other cognitive functions.